DESCRIPTION: (Applicant's Description) Germ cell tumors have been held up as a model of a curable malignancy. In addition to effective chemotherapy, successful management of this tumor has come to depend on careful definition of prognosis based on available clinical parameters. Despite a great deal of effort, the utility of these clinically-based prognostic systems has been maximized and still are associated with a 10-50 percent error rate in correctly predicting patient outcome. Newly available technologies have led to discoveries of biological and molecular events in cancer initiation, growth and metastasis, but only rarely have these discoveries been coupled to clinical outcome or led to the discovery of therapeutic targets in solid tumors. The ability to identify circulating solid tumor cells and explore genetic alterations in germ cell tumors affords the opportunity to correlate these biologic and molecular insights into issues of clinical care. We hypothesize that these new technologies can identify biological endpoints that can substantially augment or even supersede currently available clinically-based prognostic systems. A more accurate prognostic system would have significant ramifications in the selection of patients for more aggressive surgeries and treatments, and sparing some patients the rigors of such treatments. We propose to analyze molecular data obtained from the three scientific Projects in this Program in conjunction with existing clinical parameters. At completion, we will integrate these new findings into current clinical prognostic models.